gmane.comp.version-control.git

rev-list --reverse --full-history --parents with path limiter disconnects history

Johannes Sixt — 2008-08-21T16:27:50

Attached is a repository (to be imported via git fast-import) that has such a history: o--o--A--B--o \ \ o--o--C-----M--o--X <- master where the A,B,C touch path P2 such that P2 is the same after B and C. Commit X touches P2 as well. This command does not give the expected result: It reports *two* initial commits and one merge; in particular, commit X is reported as initial commit. $ git rev-list --parents --reverse --full-history master -- P2 This command gives the expected result, i.e. there is one initial commit, one merge, and some more regular commits: $ git rev-list --parents --full-history master -- P2 And this works as well, i.e there is a linear history that omits the branch via C: $ git rev-list --parents --reverse master -- P2 Please help! BTW, I've observed this behavior earlier when I debugged David Tweed's repository, after which we removed --full-history from git-filter-branch. But I could not easily find a simple history that exposes the problem. Now I have such a his

[PATCH] filter-branch: Grok special characters in tag names

Johannes Sixt — 2008-08-21T14:45:11

The tag rewriting code used a 'sed' expression to substitute the new tag name into the corresponding field of the annotated tag object. But this is problematic if the tag name contains special characters. In particular, if the tag name contained a slash, then the 'sed' expression had a syntax error. We now protect against this by using 'printf' to assemble the tag header. Signed-off-by: Johannes Sixt telecom.at> --- git-filter-branch.sh | 12 +++++++----- t/t7003-filter-branch.sh | 8 ++++++++ 2 files changed, 15 insertions(+), 5 deletions(-) diff --git a/git-filter-branch.sh b/git-filter-branch.sh index 2688254..81392ad 100755 --- a/git-filter-branch.sh +++ b/git-filter-branch.sh < at >< at > -412,15 +412,17 < at >< at > if [ "$filter_tag_name" ]; then echo "$ref -> $new_ref ($sha1 -> $new_sha1)" if [ "$type" = "tag" ]; then -new_sha1=$(git cat-file tag "$ref" | +new_sha1=$( ( printf 'object %s\ntype commit\ntag %s\n' \ +"$new_sha1" "$new_ref" +git cat-file tag "$ref" | s

Git Error

Brian Loomis — 2008-08-21T13:57:47

I'm getting the error trying to do a pull from a repos I set up on GitHub yesterday. error: some local refs could not be updated; try running 'git remote prune origin' to remove any old, conflicting branches I've run the 'git remote prune origin' but still get the error doing the pull. Any ideas how to resolve this? Google seems to show me nothing on this and I have also looked at the manual. TIA Brian Loomis http://www.virtualrelations.us (208) 639-2569 - 208 NEW BLOX

Git Error for local refs

Brian Loomis — 2008-08-21T13:58:12

Git Errors for local refs

Brian Loomis — 2008-08-21T14:12:42

I'm getting the error trying to do a pull from a repos I set up on GitHub yesterday. error: some local refs could not be updated; try running 'git remote prune origin' to remove any old, conflicting branches I've run the 'git remote prune origin' but still get the error doing the pull. Any ideas how to resolve this? Google seems to show me nothing on this and I have also looked at the manual. I've run the git fsck and git gc but still seem to get these. TIA Brian Loomis

[PATCH] add --no-verify to git-merge and git-pull

Paolo Bonzini — 2008-08-21T13:00:34

This patch adds the --no-verify option to skip the pre-merge hook. The corresponding testcase is also updated to test the new option. Signed-off-by: Paolo Bonzini gnu.org> --- Documentation/git-merge.txt | 2 +- Documentation/merge-options.txt | 3 +++ builtin-merge.c | 15 +++++++++------ git-pull.sh | 9 ++++++--- t/t5407-pre-merge-hook.sh | 15 ++++++++++----- 5 files changed, 29 insertions(+), 15 deletions(-) I'm submitting this separately because it touches a substantially disjoint set of files (besides builtin-merge.c). As an aside, I'm not particularly attached to this set of patches as long as there is a way to achieve what I wanted, i.e. blocking merge commits. One way would be with branch..mergeoptions and the fast forward strategies patch from Sverre. diff --git a/Documentation/git-merge.txt b/Documentation/git-merge.txt index 17a15ac..5a0f398 100644 --- a/Documentation/git-merge.txt +++ b/Documentation/git-merg

[PATCH] add a pre-merge hook

Paolo Bonzini — 2008-08-21T12:09:42

This hook provides a way to look at what kind of merge is invoked, and stop it. The type of merge is passed to the hook. The patch provides a sample implementation that implements a `branch..allowmerges` key. It can be useful for project whose policy is "only fast forwards on the integrator's repository". Signed-off-by: Paolo Bonzini gnu.org> --- Documentation/githooks.txt | 18 +++++ builtin-merge.c | 65 +++++++++++++++---- t/t5407-pre-merge-hook.sh | 130 +++++++++++++++++++++++++++++++++++++ templates/hooks--pre-merge.sample | 49 ++++++++++++++ 4 files changed, 250 insertions(+), 12 deletions(-) create mode 100755 t/t5407-pre-merge-hook.sh create mode 100755 templates/hooks--pre-merge.sample The included testcases demonstrate that trivial merges are currently broken. The failing test is equivalent to: git init echo a > a git add a git commit -ma git checkout -b branch echo b > b git add b git commit -mb

[PATCH] provide more errors for the "merge into empty head" case

Paolo Bonzini — 2008-08-21T12:14:18

--squash merges are not supported yet. They could be implemented by building the index from the merged-from branch, and doing a single commit. Non-fast-forward merges instead do not make sense, because you cannot make a merge commit if you don't have a base in the beginning. Signed-off-by: Paolo Bonzini gnu.org> --- builtin-merge.c | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/builtin-merge.c b/builtin-merge.c index b4c7eda..a3b9b10 100644 --- a/builtin-merge.c +++ b/builtin-merge.c < at >< at > -869,6 +869,11 < at >< at > int cmd_merge(int argc, const char **argv, const char *prefix) if (argc != 1) die("Can merge only exactly one commit into " "empty head"); +if (squash) +die("Squash commit into empty head not supported yet"); +if (!allow_fast_forward) +die("Non-fast-forward commit does not make sense into " + "an empty head"); remote_head = peel_to_type(argv[0], 0, NULL, OBJ_COMMIT); if (!remote_head) die("%s - not something we can merge", argv[

Bug/problem with 1.6.0 on UnixWare

Boyd Lynn Gerber — 2008-08-21T11:25:53

Hello, I am having the same problem on UnixWare as I have with the SCO OpenServer 6. trying to build git-1.6.0. Here is the error. CC shell.o AR compat/lib.a LINK git-shell Undefined first referenced symbol in file hexval_table abspath.o null_sha1 abspath.o trust_executable_bit abspath.o has_symlinks abspath.o UX:ld: ERROR: Symbol referencing errors. No output written to git-shell gmake: *** [git-shell] Error 1 I need some assistance to track this down. I did not have any error with the tests I did during the testing. I did not test often on all OS's. Thanks, -- Boyd Gerber zenez.com> ZENEZ1042 East Fort Union #135, Midvale Utah 84047

Bug/problem compile problem with 1.6.0 pn SCO OpenServer 6.0

Boyd Lynn Gerber — 2008-08-21T11:00:52

Hello, I finally found the time to attempt to build 1.6.0. This is what I am seeing. ... CC shell.o AR compat/lib.a LINK git-shell Undefined first referenced symbol in file hexval_table abspath.o null_sha1 abspath.o trust_executable_bit abspath.o has_symlinks abspath.o UX:ld: ERROR: Symbol referencing errors. No output written to git-shell gmake: *** [git-shell] Error 1 I did do some testing prior with my git clone of 1.6.0 and I never saw this problem. Where do I need to start to fix this. -- Boyd Gerber zenez.com> ZENEZ1042 East Fort Union #135, Midvale Utah 84047

[JGIT PATCH] Typo in implementation of pack protocol

Mike Ralphson — 2008-08-21T10:36:11

cgit's receive-pack.c has never had the corresponding typo. Fix up method and member names to match. Signed-off-by: Mike Ralphson abacus.co.uk> --- .../spearce/jgit/transport/BasePackConnection.java | 2 +- .../jgit/transport/BasePackPushConnection.java | 6 +++--- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/org.spearce.jgit/src/org/spearce/jgit/transport/BasePackConnection.java b/org.spearce.jgit/src/org/spearce/jgit/transport/BasePackConnection.java index 52f3f48..14fffc3 100644 --- a/org.spearce.jgit/src/org/spearce/jgit/transport/BasePackConnection.java +++ b/org.spearce.jgit/src/org/spearce/jgit/transport/BasePackConnection.java < at >< at > -144,7 +144,7 < at >< at > for (String c : line.substring(nul + 1).split(" ")) line = line.substring(0, nul); } -if (line.equals("capabilties^{}")) { +if (line.equals("capabilities^{}")) { // special line from git-receive-pack to show // capabilities when there are no refs to advertise continue; diff --git a/or

[JGIT PATCH] Allow make_jgit.sh compilation under Cygwin or Mingw (MsysGit) with Windows JDK

Mike Ralphson — 2008-08-21T09:33:11

Signed-off-by: Mike Ralphson abacus.co.uk> --- make_jgit.sh | 24 +++++++++++++++++------- 1 files changed, 17 insertions(+), 7 deletions(-) diff --git a/make_jgit.sh b/make_jgit.sh index 13d0e32..c119202 100755 --- a/make_jgit.sh +++ b/make_jgit.sh < at >< at > -10,14 +10,24 < at >< at > JARS=" org.spearce.jgit.pgm/lib/args4j-2.0.9.jar " -if [ -n "$JAVA_HOME" ] -then -PATH=$JAVA_HOME/bin:$PATH -fi - +PSEP=":" T=".temp$$.$O" T_MF="$T.MF" R=`pwd` +if [ "$OSTYPE" = "cygwin" ] +then +R=`cygpath -m $R` +PSEP=";" +fi +if [ "$MSYSTEM" = "MINGW" ] +then +PSEP=";" +fi + +if [ -n "$JAVA_HOME" ] +then +PATH=${JAVA_HOME}/bin${PSEP}${PATH} +fi cleanup_bin() { rm -f $T $O+ $T_MF < at >< at > -44,7 +54,7 < at >< at > do then CLASSPATH="$R/$j" else -CLASSPATH="$CLASSPATH:$R/$j" +CLASSPATH="${CLASSPATH}${PSEP}$R/$j" fi done export CLASSPATH < at >< at > -61,7 +71,7 < at >< at > do -encoding UTF-8 \ -g \ -d ../bin2) || die "Building $p failed." -CLASSPATH="$CLASSPATH:$R/$p/bin2" +CLASSPATH="${CLASSPATH}${PSEP}$R/$p/bin2" done

[PATCH] allow user aliases for the --author parameter

Michael J Gruber — 2008-08-21T09:19:41

This allows the use of author abbreviations when specifying commit authors via the --author option to git commit. "--author=$key" is resolved by looking up "user.$key.name" and "user.$key.email" in the config. Signed-off-by: Michael J Gruber fastmail.fm> --- In an ideal word, all my collaborators would exchange changes as git patches (or even via pull/push). In the real world, they send new versions which I integrate (after dealing with their whitespace and encoding changes...). Therefore, being able to say "git commit --author=mickey" and having git translate "mickey" into "Mickey Mouse ducktown.us>" is a real time saver. The patch accomplishes this by reading config keys "user.mickey.name" and "user.mickey.email" when encountering an --author argument without "<>". If there's interest in this patch I'll follow up with a documentation patch. The "--committer" argument to git commit is not treated because I don't consider it worthwhile. Note that the implementation is dif

0001-Detailed-tools-useful-when-resolving-merge-conflicts.patch

Dan Hensgen — 2008-08-21T03:27:32

Sorry, had trouble with git send-email. Thanks, Dan

[JGit Patch v2 1/7] Add a POM file for setting JGit library as a Maven project

2008-08-21T03:13:08

From: Imran M Yousuf smartitengineering.com> Signed-off-by: Imran M Yousuf smartitengineering.com> --- jgit-maven/jgit/pom.xml | 169 +++++++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 169 insertions(+), 0 deletions(-) create mode 100644 jgit-maven/jgit/pom.xml diff --git a/jgit-maven/jgit/pom.xml b/jgit-maven/jgit/pom.xml new file mode 100644 index 0000000..c4d7c24 --- /dev/null +++ b/jgit-maven/jgit/pom.xml < at >< at > -0,0 +1,169 < at >< at > + +